On Feb 24 2015 or thereabouts, Nicolas Chauvet wrote: > MSM_IOMMU doesn't play nice in multi-platform kernel > This is causing probe issue as it tries to register unconditionally > > Specially experienced on tegra_drm where IOMMU support was enabled > for tegra30 and later since 3.19. It has prevented the fallback > to non-iommu in the tegra20 case. > > This need additionals "workaround" patches for EXYNOS, OMAP and Rockchip. > v2 for thoses patches are reported at: > http://lists.linuxfoundation.org/pipermail/iommu/2015-February/012226.html > > This config change and thoses additionals patches would need > to be backported down to 3.19 fedora kernels fwiw, there was a patch upstream to mark MSM_IOMMU as depends on BROKEN.. not sure if that would be sufficient (either way, I don't really mind, the current upstream msm-iommu is not actually used by any driver, at least not until I get time to rename qcom-iommu and rework it to fit the one-true-iommu-bindings..) BR, -R > --- > config-armv7 | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/config-armv7 b/config-armv7 > index 7f76e38..4e62e3a 100644 > --- a/config-armv7 > +++ b/config-armv7 > @@ -356,7 +356,7 @@ CONFIG_PHY_QCOM_APQ8064_SATA=m > CONFIG_PHY_QCOM_IPQ806X_SATA=m > CONFIG_USB_DWC3_QCOM=m > CONFIG_CRYPTO_DEV_QCE=m > -CONFIG_MSM_IOMMU=y > +# CONFIG_MSM_IOMMU is not set > CONFIG_DRM_MSM=m > CONFIG_DRM_MSM_FBDEV=y > CONFIG_USB_EHCI_MSM=m > -- > 1.7.2.1 > _______________________________________________ kernel mailing list kernel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/kernel
